Goldman Economist: AI Productivity Gains May Take 15 Years

Goldman Sachs economist Elsie Peng has offered a cautionary perspective on the timeline for artificial intelligence to significantly impact economic productivity. In a research note published this month, Peng drew parallels between the current AI revolution and the advent of personal computers, suggesting that substantial productivity gains may not be evident for up to 15 years. Peng's analysis indicates that the computer revolution, despite initial hype, experienced a "J-curve" effect. For the first four years after the commercialization of PCs in 1981, productivity saw a modest drag. Gains only became statistically significant after eight years, and the widespread productivity boom associated with personal computers did not manifest in macroeconomic data until approximately 15 years post-commercialization.

Peng's findings highlight that the adoption and integration of new technologies, even transformative ones like AI, are heavily influenced by human factors rather than solely by technological advancement. While Goldman Sachs' official stance remains that AI will "meaningfully boost productivity growth over the next decade," Peng's work suggests that investors and proponents may be miscalculating the timing of these economic benefits. The current adoption of AI tools is described as "begrudging" and "optional" for many workers, with early use often feeling like an additional chore rather than a seamless enhancement.

The research points to the period between the early 1980s and the late 1990s as a time when investment in information and communications technology surged, yet the productivity trend remained flat. Peng's industry-panel analysis identified a four-year period of worsening performance, followed by four years of stagnation, before any measurable improvements became apparent. This historical precedent suggests that the path to realizing AI's full economic potential may involve a similar extended period of adjustment and integration, challenging immediate expectations of rapid transformation.

Peng's perspective contrasts with some of the more optimistic forecasts that anticipate AI rapidly reshaping the labor market and driving immediate economic growth. Her emphasis on the human element in technology adoption underscores the complexities involved in integrating new tools into existing workflows and organizational structures. The analogy to the computer revolution implies that the initial phase of AI integration will likely involve significant learning curves and adaptation challenges for individuals and businesses alike, delaying the widespread, measurable economic benefits.